An Israeli startup called L1ght that has created what it calls “anti-toxicity” software for online platforms said today it has raised $15 million in a seed funding round.

The round was led by Mangrove Capital Partners and Tribeca Venture Partners, with participation from Western Technology Investment.

L1ght, which was formerly known as AntiToxin Technologies Inc., has built an artificial intelligence-based platform for social media networks, messaging app providers and games platforms that helps them to identify and eradicate online bullying, hate speech, predatory behavior and harmful content. The technology works by detecting and predicting what the company calls “toxic content” in the form of text, audio, videos and images.

L1ght co-founder and Chief Executive Officer Zohar Levkovitz (pictured) said he created the anti-toxicity platform to tackle some of the dangers he felt his children face online. He said not enough was being done to address such problems.

“Our main priority has always been to keep kids safe and this investment is further proof that our technology can accomplish that,” he said.

L1ght’s technology integrates with gaming platforms, social media and messaging apps, and uses AI algorithms to privately and securely scan conversations for any evidence of abuse, bullying or grooming. It can flag a large number of different violations and then let the client decide whether or not to delete that activity, suspend the user responsible or take some other kind of action.

Because it’s based on AI, L1ght’s software can learn over time and get better at identifying whether certain behavior or content is malicious.

“The ability to place key words in the context of a conversation or online interaction is a central part of what makes our AI more reliable and sophisticated than other technologies currently available,” Levkovitz said in an interview. “Our technology helps us focus on the context of conversations rather than specific words, to try to conclude if the conversation is heading in a toxic direction.”

Levkovitz said improving these predictive capabilities was a key area of focus for L1ght as it tries to refine its technology.

“Our ambitious first goal was to prove that machine learning can be a much more effective method in identifying toxicity than today’s common practice of dictionary-based analysis,” he said. “Now, we will aim to expand and enhance our capabilities of prediction. We want to be able to stop harmful acts in their tracks, before any harm can be done to children.”

Photo: L1ght/Facebook